Since the early 1990s, many countries in sub-Saharan Africa have made significant progress in opening their economies to external competition through trade and currency liberalization. This paper analyzes trade and policy developments for 22 countries in eastern and southern Africa, looks at regional and multilateral integration issues, and reflects on the main challenges these countries face in the new decade. It addresses the main trade policy issues for these countries and suggests possible actions they and their trading partners could follow.


Throughout this paper, the ESA region will comprise the following 22 countries: Angola, Botswana, Burundi, Comoros, Democratic Republic of Congo, Eritrea, Ethiopia, Kenya, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Rwanda, Seychelles, South Africa, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe.


Middle-income countries are defined by the World Bank as those with a 1998 per capita GNP of between US$760 and US$9,360.


The trade restrictiveness index is a modified version of the one used in Sharer and others (1998), and is based on a classification system for nontariff barriers and for import and export taxes (sec Appendix I).


For example, in the case of South Africa and its partners in SACU, the maximum tariff was 1,389 percent, there were more than 13,000 tariff lines, and the effective protection ranged from—411 to 189 percent. While extreme. South Africa’s case was not atypical, as Malawi, Mauritius, and Ethiopia also had highly distorted tariff structures.


Of the 22 countries in the ESA region, 14 had programs with the IMF during the 1990s. The SACU countries (excluding Lesotho), Angola, Eritrea, Mauritius, and the Seychelles did not have programs with the Fund (see Appendix II, Table A2).


While some countries—Zambia, for example—embarked on a liberalization program that encompassed both tariff and nontariff barriers, some of the other reformers, such as Tanzania, gave priority to the elimination of nontariff barriers and reduced tariffs at a later stage.


Tentative evidence derived from developments in tariff collection ratios and a categorization of countries on the basis of the Sachs-Warner criterion suggest that the gap between ESA countries and the rest of the world has narrowed considerably since the late 1980s (Appendix II, Table A3).


Specific revisions to tariff schedules may have prevented reductions in effective protection for important sectors of the economy. In the case of South Africa, the top rate of effective protection increased from 189 percent at the beginning of the decade to 204 percent in 1998, despite the overall trade liberalization achieved during this period.


With the notable exception of the SACU, which had applied 72 tariff bands by end-1998.


Zambia and Mozambique have also moved closer to Article VIII status in 1999. However, capital account liberalization has not featured prominently in the reform agenda of most countries in the region. Except for Mauritius, Uganda, and Zambia, several capital account controls are still in place in most countries.


Including exemptions on imports by nongovernmental organizations and those financed by foreign donors, which usually cover a large part of total imports.


In Mauritius, the share of exempted imports in total imports increased by about 10 percent during the 1990s. In Tanzania, the scope of exemptions has also increased in recent years, as evidenced by the rise in the ratio of forgone revenue relating to exemptions to total import taxes assessed (including exemptions), from 24 percent in 1993 to 48 percent in 1997–98. In Mozambique and Zambia, the authorities have emphasized recently the use of discretionary exemptions in an attempt to promote foreign investment.


See, for instance, Rodrik (1998).


Zambia introduced an across-the-board 5 percent import declaration fee in 1995 for fiscal reasons, which was eliminated in 1998. Also, an import ban on wheat flour, introduced in 1997, was lifted in the following year. The 10 percent export levy on tobacco, tea, and sugar introduced by Malawi in 1995 for fiscal reasons was eliminated in 1998. In Tanzania, some of the tariff increases in creases in 1992 were reversed during See. for instance, Rodrik (1998).


Binding represents a legal commitment by a country not to raise tariffs above a specified level. If a binding is breached by a country it has to compensate partner countries or face retaliation by them.


Computed as Hi, = Σj=1,…S(Sjj=1,….SSj)2, where Sj is the sum of deposits and loans for bank j. As defined, the index equals one in the case of a monopoly and 0 in the case of equal shares. An index value higher than 0.3 is normally associated with an imperfect market structure.


The liberalization index corresponds to the bound rather than the applied level of access, and, therefore, does not necessarily correlate with how liberal actual conditions are.


Tanzaiiia will withdraw from COMESA effective September 2000.


Angola, the Democratic Republic of Congo, and the Seychelles are the only three countries that are not signatories of the protocol, though Angola and Seychelles have signaled Their intention to accede to it in the near future.


ESA countries that are net importers of foodstuffs could also be affected by the increase in prices brought about by the reduction in subsidization of agricultural products. This increase is, however, expected to be small.


These preferences stem from various arrangements between industrial countries and ESA countries, such as the Generalized System of Preferences between the European Union. Japan, and the United States, on the one hand, and developing countries, on the other. There are also the Lomé Convention, involving the EU and several low-income countries, including ESA countries, and others, such as the Multi-Fiber Arrangement, under which several ESA countries have special quotas for the export of textiles and clothing products.


Tariff escalation is a phenomenon whereby tariffs increase with the stage of processing.


This result is based on Yeats (1998) and covers eight ESA countries for which reliable data are available at a disaggregated level. The countries are Kenya, Madagascar, Malawi, Mauritius, South Africa, Uganda, Zambia, and Zimbabwe. The non-ESA sub-Saharan African countries covered are Benin, Burkina Faso, Cameroon, Gabon, Côte d’lvoire, and Senegal.


According to Amjadi and Yeats (1995), net payments on freight and insurance by sub-Saharan African countries were in 1990–91 equivalent to about 15 percent of exports of goods and services, more than twice as high as in other developing countries. In landlocked countries such as Malawi and Uganda, the ratio was well above 50 percent.


The single-product and three-product concentration ratios measure the share of total exports accounted for by the largest and the three largest export items, respectively.


The following variables are analyzed: GDP growth, GDP per capita growth, growth in export volume, inflation performance, the fiscal balances, the current account balance (the latter two excluding grants and expressed as a share of GDP), and indicators of trade revenue dependency. The Democratic Republic of Congo and Eritrea are excluded from the analysis because their macroeconomic data are either unavailable or highly distorted by the recent wars in these countries.


In virtually all cases, those countries that had open trade regimes in 1998 were also the ones that undertook the most liberalization since the beginning of the decade.


The BLNS countries are excluded from the analysis here because their “trade taxes” largely represent transfers from South Africa under the revenue-sharing arrangements in SACU.


The African Growth and Opportunity Act, enacted by the U.S. Congress in early 2000, provides for duty-free treatment of the exports of sub-Saharan African countries to the United States. The magnitude of benefits to these countries, especially in relation to textiles and clothing products, will be determined by rules of origin requirements that need to be fulfilled to qualify for duty-free access. An interesting feature of the Act is that it envisages the possible negotiation of free trade agreements between the United States and African countries in the future.


Tariff escalation is significant for many product chains relevant to sub-Saharan African countries, such as wood products, textiles and clothing, fish, and leather products (Amjadi, Reinke, and Yeats, 1996).


Pre-Uruguay Round tariffs facing sub-Saharan African countries’ exports to the EU. United Stales, and Japan averaged only 0.31 percent, less than for any other group of developing countries (Amjadi, Reinke, and Yeats. 1996).

Appendix I Overall Trade Restrictiveness Classification Scheme

As noted in Sharer and others (1998), the restrictiveness of the tariff regime depends on many factors, including the minimum and maximum tariff rates, the number of bands, the allocation of individual items to the bands, the existence of “exceptional” rates that lie outside the basic tariff structure, any other duties and charges (such as differential rates of excise or value-added taxes on imports, import surcharges, and statistical fees), and the extent of exemptions from customs duties. The classification scheme for overall trade restrictiveness used in this study builds on the approach developed by Sharer and others by including information on export taxes and by adding another category of nontariff barrier (NTB) restrictiveness.

Under this approach, a country is assigned a restrictiveness ranking for both its price-based (import tariffs and export taxes) measures and its NTBs, based on the classification schemes described below. Consistent with the Lerner theorem that an export tax is equivalent to an import tax, the price-based measure is computed as the sum of import and export taxes. Whenever possible, an unweighted average of statutory tariff rates including other duties and charges was used. An average of statutory tariff rates is preferable to one based on customs collections since the latter reflects (often extensive) exemptions. An unweighted average is preferable to a trade-weighted average since items with high tariffs are likely to have small trade weights. Other duties and charges should also be included. The restrictiveness ranking accords greater weight to NTBs, which are inherently less transparent and more distortionary than tariffs.

NTBs include quantitative restrictions, restructure licensing, bans, state trading monopolies, restrictive foreign exchange practices that affect the trade regime (for example, a surrender requirement at a non-market exchange rate, multiple exchange rates), quality controls, and customs procedures that act as trade restrictions. Such measures in effect provide indirect subsidies to import-competing domestic producers in a nontransparent manner. However, information on NTBs and their restrictiveness as measured, for instance, by ad valorem equivalents.1 may either not be available for all countries or be of limited use,2 A review of previous studies on trade reform in developing countries below shows that other researchers have faced similar difficulties. In view of this, this study utilizes four categories of NTBs (see table above).

Overall Trade Restrictiveness Classification Scheme

article image

Includes restrictions on exports and imports and other NTBs.

includes customs duties and all other charges levied exclusively on imports, as well as export taxes.

Refers to the share of total trade being affected by NTBs.

The table illustrates the assignment of price-based and NTB categories on the 10-point scale. A country is assigned an overall rating of 10 as long as the average tariff rate exceeds 35 percent, even if the nontariff trade regime is classified as open. Countries with an overall rating of between 7 and 10 are classified as restrictive; those with a 5 or a 6 rating are classified as moderately restrictive; those with 3 or 4 as moderately open; and those with 1 or 2 as open.


The ad valorem equivalent of an import quota is the rate of advalorem tariff that would yield the same import quantity as the quota. There are many circumstances in which import quotas and import tariffs are not equivalent, including imperfect market structures and changes in supply and demand.


Trade or production coverage of NTBs may be useful but does not fully capture their restrictiveness.

Appendix II Statistical Tables

Table A1

Eastern and Southern Africa: Structure of Production (Constant Prices), 1985–971

article image
Source: World Bank, World Development Indicators, 1999.

Ratio of appropriate sectoral variable to GDP at factor cost. All averages are unweighted unless indicated otherwise. In a few cases, shares do not add up to 100.

Final period refers to 1995.

Initial period refers to 1988.

Initial period refers to 1988 and final period to 1996.

Initial period refers to 1988.

Table A2.

Eastern and Southern Africa: Countries with IMF Arrangements During the 1990s

article image
Source: IMF, international Financial Statistics.
Table A3.

Sachs-Warner Classification of Trade Policy of ESA Countries, 1980s and 1990s

article image

IMF staff’s calculations applying the Sachs-Warner criterion for tariffs and nontariff barriers. That is, a country is classified as closed if its NTBs covered 40 percent or more of the value of trade or if its average tariff exceeded 40 percent.

Table A4

Eastern and Southern Africa: Current Account Exchange Restrictions, 1990 and 1999

article image
Sources: IMF (1991); and IMF (2000).

As of December 31.

As of December 31, 1999, in percent of total.

Table A5

Eastern and Southern Africa: Nontariff Barriers to Imports, December 1998

article image
Sources: World Bank and IMF staff reports; and data provided by the authorities.

Includes countervailing duties, dumping, etc.

Table A6

Eastern and Southern Africa: Nontariff Barriers to Exports, December 1998

article image
Sources: World Bank and IMF staff reports; and data provided by the authorities.

Only for restrictive (and not for statistical) purposes.

Except for diamonds.

All exports, except to SACU member countries, require a license. Within SACU, textiles and meat products require a license.

Table A7

Eastern and Southern Africa: Summary of Uruguay Round Commitments in Agriculture and Industry

article image
Sources: Sorsa (1996); and IMF staff estimates.

Other duties and charges.

Table A8

Eastern and Southern Africa: Commitments Undertaken in Trade in Services in the WTO

article image
Source: Based on schedules of commitments submitted to the WTO.
Table A9

SADC: Intraregional and Extraregional Trade

article image
Source: IMF Direction of Trade Statistics

Import and export data differ mainly because partner country data are used directly or indirectly in estimating missing figures. Imports are measured c.i.f., exports are measured f.o.b.

Table A10

COMESA: Intraregional and Extraregional Trade

article image
Source: IMF, Direction of Trade Statistics.

Import and export data differ mainly because partner country data are used directly or indirectly in estimating missing figures. Imports are measured c.i.f., exports are measured f.o.b.

Table A11

CBI: Intraregional and Extraregional Trade

article image
Source: IMF, Direction of Trade Statistics

Import and export data differ mainly because partner country data are used directly or indirectly in estimating missing figures. Imports are measured c.i.f., exports are measured f.o.b.

Table A12

Selected Eastern and Southern African Countries: Tariff Preferences

(In percent, unless otherwise indicated)

article image
Sources: Yeats (1994); and Harrold (1995).

Difference between unweighted tariff faced by African exports and those paid by competitor countries

Table A13

Sub-Saharan Africa: Compound Annual Growth Rates for Exports in U.S. Dollars


article image
Source: Yeats (1998).